I have looked for many years to locate a recording ammeter that would allow multiple channels of current flows to be recorded simultaneously to a computer and cost very little to implement. They are commercially available, but those I find are very high priced.<p>I do not need super accuracy or high speed data refresh rates. The purpose is to monitor the current used in my home circuits continuously to plot an itemized estimate of which circuits are heavily used over time and how changes in my usage changes with time or habits.<p>I propose putting small (“secondary” coils of wire on each A/C wire coming from each breaker in my house. This would convert the A/C amps to a voltage that could go to a A/D converter for each channel (circuit). If one A/D converter could be “multitasked” so 20 or more channels could cycle through the converter this may keep the hardware cost down. Having the raw data go into either a .CSV text file or directly into Excel would be great. Calibration could be “seat-of-the-pants”-- say by using a single clamp-on type Amprobe multimeter to calibrate the current on each channel to “set”.<p>Anyone having ideas on plans already available or idea on how to do this project CHEAPLY would be appreciated. Other ideas welcome too!<p>Thanks,
Charles
I'm glad you want to monitor AC current because DC is a pain in the neck unless you use a series resistor.<p>What you want is a simple data acquisition system. Almost any system with a few channels of A/D would work. Check out this URL http://www.dataq.com/products/startkit/di154rs.htm they have a 4 channel kit for $150 but more importantly a starters kit ( they used to give a free sample to qualified OEM clients) for $30 so you can develop the sensor and software without spending too much. If you put a diode and series resitor across the output of the coil you should be able to sense the current with a differential amplifier attached to both sides of the resistor. The gain and resistor value depend on the number of turns in the sense coil, and the sensitivity and range of the current you want to measure. Some data Acq cards have differnntial inputs you can use for this directly. <p>You may find it easiest to calibrate it by manually setting test currents and recording the A/D output then supply the program with a lookup table (A/D counts = I) instead of trying to figure out some sort of formula to characterize your sensor.<p>You could probably make your own system from scratch. All you need is an A/D chip and a shift register to get the output onto the serial port. Check out http://www.maxim-ic.com/appnotes10.cfm/filter/category they have lots of application notes. Look at the Max 1452 serial comm chip and measurement circuits category<p>They have off the shelf products for monitoring energy usage that cost in the range of $20 to $40 per installation. Monitoring, analysis of usage and adjustmant of appliances is what the future really holds for the internet refrigerator.
Here is the absolute minimum solution i can think of.It uses one differential op amp pre channel and a basic stamp.
Use the pwm output of a basic stamp to feed the referance input of all of the op amps The input from your current transformers will go to the other inputs of the op am. Use the remaining stamp outputs to multiplex the op amp outputs into one stamp input.
In software you select a channel and then read it by ramping the pwm reference output up until it exceeds the peak input and cuts off the pulses which will be coming from the amp.
Of course this will take some development but you should be able to do it with cheap amps and logic circuits and lot of experimenting.
Their (Dataq) DI-194RS model has 10bit resolution and may be just fine for my hobby application.<p>Any idea how to expand these to 16 channels short of putting 4 serial ports in my PC?<p>Thanks for the ideas so far!!!<p>Charles
coils of wire on each A/C wire coming from each breaker in my house. This would convert the A/C amps to a voltage that could go to a A/D converter for each channel (circuit). If one A/D converter could be “multitasked” so 20 or more channels could cycle through the converter this may keep the hardware cost down. Having the raw data go into either a .CSV text file or directly into Excel would be great. Calibration could be “seat-of-the-pants”-- say by using a single clamp-on type Amprobe multimeter to calibrate the current on each channel to “set”.<p>Anyone having ideas on plans already available or idea on how to do this project CHEAPLY would be appreciated. Other ideas welcome too!<p>Thanks,